1. Field of the Invention
The present invention relates to the field of image processing systems, and more particularly, to conversion of motion picture film to a high definition video format.
2. Art Background
Telecine systems have been developed for converting motion picture film images into conventional NTSC and PAL video signal formats. Traditionally, these telecine systems were developed to convert the motion picture film images to television signals for broadcast. Prior to the emergence of solid state imagers, telecine equipment was constructed using camera tubes or flying-spot scanners. The flying-spot scanner generates a video signal from film by scanning the film image with a very small spot of light and collecting the resulting transmitted light at a photo cell. Although an effective means to convert film to NTSC or PAL video signal formats, flying spot scanners generate video signals with limited resolution. Therefore, flying spot scanners are not an effective means for film to high definition video conversion. As equipment such as telecines are developed to facilitate the interface between high definition video and film, it is important to address all aspects of the system interface to provide quality film to high definition video conversion.
In general, telecine systems generate light to create optical film images from devices known as lamphouses. Traditionally, lamphouses for telecines and optical printers utilize specular light sources. The specular light sources, in conjunction with reflector and condensing lens types, are typified by high efficiency and contrast. Although specular light sources and the associated optics exhibit high efficiency and contrast, the specular light sources require critical optical alignment of lamp filament, reflector and condensing lenses. This critical alignment also varies with magnification further emphasizing the need for precise optical alignment. Additionally, a specular source increases the visibility of surface abrasions on the film.
The color television cameras used in traditional telecine systems typically have spectral sensitivities and pre-amplifier gain characteristics optimized for a 2900.degree. K. to 3200.degree. K. light source. These light sources, such as tungsten light sources often used in television studios, are appropriate for the spectral sensitivities of the Vidicon type detectors. However, a high definition charged coupled device (CCD) camera has different spectral characteristics than the Vidicon type detectors. The spectral sensitivity of present CCD detectors is characterized by poor blue sensitivity and excellent red sensitivity into the infra-red region. Therefore, in telecine systems utilizing CCD detectors, it is desirable to provide a light source in the lamphouse that provides efficient alternatives to traditional tungsten light sources by exhibiting spectral bandwidths and color temperatures maximized for the CCD detector.
A television camera often contains an optic prism to separate the input polychromatic optical image into its three constituent monochromatic components of red, green and blue (RGB). Existing color television camera prism designs make a very efficient use of the available light spectrum. Typical prism designs result in greater than 80% of the spectral bandpass being directed to the appropriate red, green or blue detector. Unfortunately, the separation into the RGB components in the prism results in a certain amount of crosstalk between the RGB detectors. In order to reduce or eliminate the effects of the crosstalk among the CCD detectors, a matrix is required in the television camera signal processing circuit. Although the matrix helps to reduce or eliminate the CCD detector crosstalk, it results in a reduction of the signal to noise performance of the system. Therefore, it is desirable to generate a light source that minimizes the crosstalk among CCD detectors so as to reduce or eliminate the need for the crosstalk matrix.
Color film emulsions suffer from a similar crosstalk problem. In fact, the problem is more severe since the dyes generated in the film emulsion have considerably inferior bandpass characteristics to that of the interference filters employed in television camera prisms. However, color negative films employ a very efficient photo-chemical matrix or "dye masking" system which results in excellent color response which enables color print film colorimetry to exceed the present display colorimetry of high definition video signals (HDVS). When color film emulsions are the originating source for HDVS images, an appropriate color separation system is required which is substantially different to that of a color television camera imaging the real world. The most important characteristic of this separation system is that it creates color separations based on the primary analysis characteristics of the film emulsion while minimizing electrical crosstalk, and therefore reducing the magnitude of matrix coefficients.
An appropriately designed lamphouse can contribute to overall system performance by providing optimal conditions to realize best signal to noise performance. Therefore, it is desirable to create a lamphouse that provides a light source designed to compliment the scanner spectral characteristics, minimize crosstalk among the color channels, and provide an even light field which can reduce shading compensation.
Conventional motion picture film consists of frame images which are commonly displayed sequentially at a rate of 24 frames per second (fps). However, the standard video frame rate is 25 video fps for PAL format video, 29.97 video fps for the NTSC video format and 30 video fps for the SMPTE-240M high definition video format. Therefore, to convert motion picture film images into video image signals, frame conversion is required. In PAL systems, it is a universal practice to reproduce the 24 frame rate film at 25 fps. In practice, this is acceptable because the discrepancy is only 4.17%, and the increase in pitch and reduction in running time of the film is acceptable. In the case of NTSC and high definition video, the discrepancy of 6 frames per second is a 25% difference, and therefore, reproducing the 24 fps motion picture film would be intolerable. To compensate for the frame rate mismatch, a technique commonly referred to as 3-2 pull down in used to generate higher video frame rates.
Typically, the 3-2 video frames are generated by holding the film in a projector gate to permit two field exposures or three field exposures to occur. A pulldown mechanism, which transports the film into the projector gate, delivers the film at uneven periods in order to generate either the two or three exposures. In addition, the pulldown mechanisms used to achieve the 3-2 film frame to video frame conversion are edge guided. The edge guided system does not provide accurate image registration. While edge guiding may provide acceptable results for NTSC and PAL video conversion, a highly accurate positioning pulldown mechanism is required for high definition video. Therefore, it is desirable to generate the 3-2 pulldown film to video frame conversion rate with the same high precision frame placement system in which the original image was created on the camera.